This invention relates to an apparatus and method for compensating component audio signals.
One example of compensating component audio signals is equalizing a soundfield. As used herein, the term "soundfield" means an environment in which an audio frequency signal is propagated in electrical form from a signal source to an electroacoustic transducer and is propagated in acoustic form to a receiver. The source of the audio frequency electrical signal may be, for example, a microphone, the cartridge of a phonograph, the tuner of an FM or AM radio or the playback head of a tape player. The electroacoustic transducer may be a loudspeaker, and the receiver may be an acoustoelectric transducer or the human auditory system. Even when the receiver is an acoustoelectric transducer, for converting the audio frequency acoustic signal to electrical form, the signal will normally be applied eventually to the auditory system of a human listener. Thus, as used herein the term "soundfield" is not limited to an environment in which actors, singers or musicians generate vocal or musical sounds to be heard immediately by, or recorded for subsequent playback to, other persons, but applies also to, for example, an environment provided with a public address system, such as an airport terminal building, and a room having a sound amplification system, such as a courtroom.
The auditory sensation received by a person listening (either directly or indirectly) to the sound emitted by the electroacoustic transducer depends not only on the signal provided by the source of the audio frequency electrical signal but also on the frequency response of the signal propagation path from the signal source to the receiver. The frequency response of the propagation path defines the relationship between the amplitude of a signal component generated at the source and the amplitude of that same signal component provided at the receiver, as a function of the frequency of the signal component. If the frequency response of the propagation path is not uniform, the relative amplitudes of the signal components at different frequencies are altered during propagation from the signal source to the receiver.
Generally, the frequency response of a signal propagation path is not uniform, and it is conventional to compensate for the non-uniformity in the frequency response of the propagation path from the signal source to the receiver by including an equalizer in the propagation path. An equalizer is an amplifier of which the frequency response is selectively variable. In general, if, in the absence of the equalizer, a signal component at a given frequency is attenuated in the propagation path to a greater degree than are signal components at other frequencies, the frequency response of the equalizer is adjusted so that the gain of the amplifier is greater at the given frequency than at other frequencies.
As shown in FIG. 1, a known form of an equalizer 2 comprises an operational amplifier 4 having a non-inverting input terminal which receives the signal from a signal source 6 by way of an audio amplifier 8. The amplifier 4 has an output terminal which is connected through a resistor 10 to an output node 12 which is in turn connected through a power amplifier (not shown) to a loudspeaker 14. A feedback resistor 16 is connected between the output terminal of the amplifier 4 and its inverting input terminal 17. The equalizer also comprises a bank of n resonators 18i (i=1 . . . n) of selected center frequencies fi. A variable resistor 20i is connected between each resonator and an associated switch 22i. In a first position of the switch 22i, the series combination of the resistor 20i and the resonator 18i is connected between the terminal 17 and ground, and in a second position of the switch 22i the series combination is connected between the node 12 and ground. Ignoring the effect of the other resonators and resistors, when the switch 22.sub.1 associated with the resonator 18.sub.1 having the center frequency f.sub.1 is in the first position, the gain A.sub.v1 at the frequency f.sub.1 is given by ##EQU1## where R16 is the resistance of the resistor 16.sub.1, R20.sub.1 is the resistance of the variable resistor 20.sub.1 and Z.sub.1 is the impedance of the resonator 18.sub.1 at the frequency f.sub.1 When the value of R20.sub.1 is infinite, A.sub.v1 is equal to unity.
When the switch 22 is in its second position, the gain A.sub.v1 is one and the amplitude of the signal applied to the loudspeaker 14 at the frequency f.sub.1 depends on the potential divider effect of the resistor 10 and the series combination of the resistor 20.sub.1 and the resonator 18.sub.1. The signal applied to the loudspeaker 14 at the frequency f.sub.1 is attenuated by a factor EQU (R20.sub.1 +Z.sub.1)/(R10+R20.sub.1 +Z.sub.1)
with respect to the signal at the output terminal of the amplifier 2. By appropriate selection of the values of the resistors 20i and the settings of the switches 22i, a frequency response curve can be established such that at certain frequencies the overall gain of the propagation path, including the equalizer, is unity, at other frequencies it is greater than unity and at still other frequencies it is less than unity.
The resonators 18 may each comprise a series combination of a resistor, an inductor and a capacitor. The impedance of the resonator is adjustable by varying the resistance value of the resistor. The resonators may be of constant Q (the value of Q is independent of the resistance value of the resistor) or of variable Q (the value of Q depends on the resistance value of the resistor).
In a conventional equalizer, the positions of the switches and the values of the variable resistors are adjusted through slide controls on a useraccessible front panel of the equalizer. The settings allow the user of the equalizer to determine at a glance the configuration of the frequency response curve of the equalizer, and also allow the user to set the curve to a desired configuration, i.e., boosting certain frequencies and cutting other frequencies. This feature of the conventional equalizer is, however, subject to misuse, particularly when the soundfield requires the services of a skilled audio technician to achieve satisfactory equalization, since an unskilled person can adjust the settings of the slide controls as easily as can a skilled technician, with the result that the equalization achieved by the technician is lost and cannot be recreated without expenditure of substantial effort.
The National Semiconductor LMC835 integrated circuit is a programmable resistor network suitable for use in conjunction with an operational amplifier and a bank of fourteen resonators to provide an equalizer. As shown in FIG. 2, the LMC835 comprises multiple resistors 30 connected in series with respective switches 34. The resistors are arranged in fourteen groups of twelve, which are connected to the resonators 18 respectively. Only one of the fourteen groups of resistors is shown in FIG. 2. Each group of twelve resistors is divided into a boost network 38 and a cut network 42. In the boost mode, all the switches connected to the resistors of the cut network 42 are non-conductive and therefore the cut network is out of circuit. One or more of the switches connected to the resistors of the boost network are conductive, and the resonator 18 is connected in series with the boost network (of which the resistance depends on which of the switches 34 are conductive) between the inverting terminal 17 of the amplifier 4 and ground. In the cut mode, all the switches connected to the resistors of the boost network 38 are non-conductive and one or more of the switches connected to the resistors of the cut network 42 are conductive.
The permitted states of the switches 34 allow twelve different boost settings, twelve different cut settings and a neutral, or unity gain setting. The states of the switches are controlled by serial data applied to the integrated circuit by way of a serial data port having data, clock and strobe lines, and a serial-to-parallel register 46. The serial data is applied to the register 46 in a succession of data sets, each of which comprises band selection data and gain selection data. The state of the data line is read on successive clock pulses, and a strobe pulse signifies the end of a data set. The strobe pulse causes the preceding data set to be latched. The band selection data is latched into a latch 48 and is decoded by a decoder 50 to select one of the fourteen banks of twelve resistors, and the gain selection data is latched into a latch 52 and is applied to a selector circuit 54. The circuit 54 responds to the gain selection data by causing selected switches 34 to be rendered conductive. Hitherto, it has been conventional to supply the serial data to the LMC835 integrated circuit by use of front panel controls on the equalizer. For example, as shown in FIG. 3, the equalizer might be provided with a cut switch 56 and a boost switch 58 for each group of resistors. On momentarily closing one of the switches of the pair, e.g. the cut switch 56, a microcomputer 60 generates a data set which is applied to the LMC835 circuit 64. The band selection data of the data set identifies the group of resistors of the circuit 64 that is associated with one of the resonators 18, and the gain selection data causes switches 34 of the circuit 64 to be closed such that the gain of the equalizer at the frequency associated with the selected group of resistors is reduced by one step.
A disadvantage of the LMC835 integrated circuit is that on loss of operating power for the equalizer, due to switching off of the equalizer or to interruption of the utility service supply, the gain settings are lost. In order to allow the frequency response curve to be recreated, an equalizer employing the LMC835 integrated circuit may include an LED bar graph display 62 which is driven by the signals provided by the microcomputer and provides a visible indication of the shape of the frequency response curve of the equalizer. After interruption of the power supply, the user of the equalizer can adjust the boost and cut switches to recreate the frequency response curve previously shown by the bar graph display, provided of course that the shape of the curve is known.
Another example of compensating component audio signals is mixing audio signals from different sources. FIG. 7 is a block diagram of an audio mixer based on known principles. Mixer 156 has six identical input channels which receive audio signals from respective audio signal sources 152A-152F. Each source 152 might be, for example, a microphone, the cartridge of a phonograph, or the playback head of a tape recorder. The signal provided by source 152A is amplified by an audio amplifier 154A. The output signal of amplifier 154A is applied to an input attenuator 170A which provides a signal of which the amplitude represents the product of the amplitude of the input signal and the attenuation setting of the attenuator. Input attenuator 170A is illustrated as a voltage divider comprising a fixed resistor 184A, a variable resistor 186A and a scaling resistor 187 connected in series between the output terminal of amplifier 154A and ground, with the output signal of attenuator 170A being taken from the connection point of the resistors 184A and 186A through an operational amplifier 188A. The output signal of attenuator 170A is summed in a summing circuit 174 with output signals of the other input attenuators 170B-170F, and summing circuit 174 outputs a single audio signal to an output attenuator 178. Output attenuator 178 is similar to input attenuator 170A and provides an output signal that is a function of the attenuation setting of attenuator 178. The signal provided by attenuator 178 is applied through an operational amplifier to an output amplifier 158. The output amplifier 158 is connected to an equalizer (not shown) whereby the frequency response of the audio signal is adjusted and then the resultant signal drives a loudspeaker (not shown) through a power amplifier (not shown).
Referring again to FIG. 2, the fourteen groups of resistors of the LMC835 circuit are in two sets, which may be designated set A and set B, and all the cut networks of set A are connected to a single pin of the circuit. Similarly, all the boost networks of set A are connected to a second pin, all the cut networks of set B are connected to a third pin, and all the boost networks of set B are connected to a fourth pin. Referring to FIG. 7, the variable resistors 186 of the attenuators are implemented using LMC835 circuits 150-153 operating under control of a microcomputer 168. In the mixer shown in FIG. 7, the variable resistor 186 of input attenuator 170A is implemented by the cut network of one of the seven groups of resistors in set A of LMC835 circuit 150. The variable resistor of the input attenuator 170B is implemented by the cut network of one of the seven groups of resistors in set B of circuit 150. Circuits 151 and 152 are connected, in similar fashion to circuit 150, as variable resistors of the other input attenuators, and the cut network of one of the seven groups of resistors in set A of circuit 153 provides the variable resistor of the output attenuator. The seven groups of resistors in set B of circuit 153 are unused. The mixer shown in FIG. 7 has a user interface 162. User interface 162 includes a cut switch 164 and a boost switch 166 for each of the six input channels A-F and for the output attenuator. The cut and boost switches are actuated by buttons on the user interface. User interface 162 operates in similar fashion to the user interface shown in FIG. 3. In response to the closing of a cut or boost switch of the user interface, microcomputer 168 generates a data set which is applied to circuits 150-153. Each data set includes not only band selection data and gain selection data but also chip selection data, for selecting which of the LMC835 circuits is to be adjusted in response to the band and gain selection data of that data set. The chip selection function may alternatively be accomplished by a decoder (not shown) which applies the band and gain selection data to only one selected LMC835 circuit.
If, for example, user interface 164 selects channel A, circuit 150 receives the channel selection data and gain selection data from microcomputer 168. The channel selection data selects the group of resistors whose cut network implements variable resistor 186A, and the gain selection data selects a resistance value for the cut network of the selected group such that the output signal of amplifier 188A has the desired amplitude relative to the signal provided by amplifier 154A. The output signal of amplifier 188A is applied to summing circuit 174. When the mixer is powered up, microcomputer 168 provides data sets such that the other six groups of resistors in each of sets A and B are all in the neutral, or unity gain, mode, and therefore have no effect on the output signal provided by attenuator 170A.
Similarly, if user interface 164 selects output attenuator 178, circuit 153 receives the channel selection data and gain selection data from microcomputer 168. The channel selection data selects the group of resistors that implements the variable resistor of attenuator 178, and the gain selection data selects an appropriate resistance value for the cut network of the selected group.
The controls of a conventional mixer may be similar to those of an equalizer, and allow the user to determine at a glance the relative contributions of the input signals to the output signal. However, the conventional mixer is subject to the same disadvantage as the conventional equalizer with regard to adjustment by an unskilled person.
Prior to Jan. 29, 1986 a computer-controlled mixer had been publicly used. This mixer has multiple input channels receiving audio signals from respective sources, a summation circuit and an output channel. The gain of each input channel and of the output channel is controlled by means of a general-purpose digital computer. The mixer may be used in a sound studio, where a sound engineer dynamically controls the gain setting of each input channel to achieve a desired aural impression. However, the mixer cannot be used as a stand-alone unit, since the gain settings supplied to the input channels are not stored in nonvolatile memory and are lost when the computer is disconnected from the mixer.